Skills infsh-cli

📦

infsh-cli

Name: infsh-cli
Author: qu-skills

Content revision r1 Critical 🌐 Network access🔑 Env variables📁 Filesystem access⚙️ External commands

Run AI Apps from the inference.sh CLI

Teams need one command path for many cloud AI models. This skill helps users find apps, prepare inputs, run jobs, and track results with the belt CLI.

Supports: Claude Codex Code(CC)

⚠️ 38 Poor

Do not auto-install this skill.

The canonical policy requires operator review before any installation action.

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request

Review the Skillstore skill "infsh-cli" from https://skillstore.io/skills/qu-skills-infsh-cli.md and its manifest at https://skillstore.io/api/skills/qu-skills-infsh-cli/manifest. Verify the artifact. Do not auto-install. Inspect the skill and report your findings, then wait for an operator or manual installation decision.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/qu-skills-infsh-cli.md Signed manifest GET /api/skills/qu-skills-infsh-cli/manifest Signed lockfile GET /api/skills/qu-skills-infsh-cli/lockfile

Test it

Using "infsh-cli". Find a fast video generation app and prepare a first run.

Expected outcome:

Suggested app: google/veo-3-1-fast for short video generation.
Next step: inspect the app details, create a sample input file, then run with your prompt.
Safety note: review cost and provider terms before submitting the job.

Using "infsh-cli". Run an upscaler on a local product image.

Expected outcome:

Selected app: falai/topaz-image-upscaler.
The local image path will be uploaded to inference.sh for processing.
I would ask for confirmation before uploading the file.

Using "infsh-cli". Check the result of a long-running generation task.

Expected outcome:

The skill can retrieve the task status, show whether it is complete, and save the resulting media link when available.

Security Audit

Critical

v5 • 7/9/2026 Open versioned report

The most serious confirmed issue is repeated pipe-to-shell installation from a remote URL, which is a critical supply-chain execution pattern. Several Markdown command examples are safe documentation, but the skill also enables remote app execution, local file uploads, and Twitter/X write actions that need explicit user confirmation.

Files scanned

608

Lines analyzed

Review items

False positives ignored

Confirmed security concerns (6)

Critical

references/authentication.md:6

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The command pipes a remote script directly into sh. If the endpoint or network path is compromised, arbitrary shell code can run on the user machine.

Critical

references/authentication.md:44

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The reinstall path repeats the remote script piped into sh. This remains a critical supply-chain execution pattern even when presented as documentation.

Critical

references/cli-reference.md:6

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The installation reference pipes a remote script into sh. This pattern can execute arbitrary code from the remote host on the local system.

Critical

SKILL.md:18

Pipe to shell pattern

curl -fsSL https://cli.inference.sh | sh

The primary install command pipes a remote script directly into sh. This is a critical remote-code-execution supply-chain pattern for a marketplace skill.

High

SKILL.md:58-73 references/running-apps.md:21-42

Local File Uploads to Cloud Apps

The skill explains that local file paths are automatically uploaded when passed to belt app commands. This can disclose private images, audio, video, or customer files if the agent runs commands without an explicit file review.

The documentation directly states that local files are automatically uploaded when paths are used, and examples include image, audio, and media paths.

High

SKILL.md:4 SKILL.md:51-52 SKILL.md:72-73 SKILL.md:99

Unconfirmed Social Media Write Actions

The skill includes Twitter/X posting and media-posting commands, and its tool permission allows belt command execution. Without a required confirmation step, an agent could publish or modify social content through connected accounts.

The allowed-tools entry permits belt commands, and multiple documented examples use X/Twitter write-capable apps.

Capability review items (19)

These are real local capabilities that may be expected for this skill, so they require review but are not counted as confirmed malicious behavior.

High

SKILL.md:33

Hidden file in home directory

> mv inferencesh-cli-* ~/.local/bin/inferencesh

The command moves a downloaded executable into ~/.local/bin so it can run from PATH. That is a legitimate install location, but it creates a persistent executable from remote artifacts.

Medium

SKILL.md:7

Ruby/shell backtick execution

> **Install the belt CLI skill:** `npx skills add belt-sh/cli`

The line instructs users to install another CLI skill through npx. It is user-visible, but it fetches external package content and expands the execution surface.

Medium

SKILL.md:17-20

Ruby/shell backtick execution

```bash

The shell block contains a remote installer command followed by login. The risk is explicit command execution from documentation, not Ruby backtick execution.

Medium

SKILL.md:25-34

Ruby/shell backtick execution

> ```bash

The manual installation block downloads artifacts, verifies a checksum, extracts files, and moves a binary into PATH. It is safer than pipe-to-shell but still a remote installation workflow.

Medium

SKILL.md:38-56

Ruby/shell backtick execution

```bash

The examples run belt app commands against remote AI services and include a social posting example. The commands are explicit, but they can create external side effects.

Medium

SKILL.md:62-74

Ruby/shell backtick execution

```bash

The shell block runs belt app commands with local file paths and a social media post with media. These commands can upload files or publish content to external services.

Medium

SKILL.md:86-87

Ruby/shell backtick execution

| Run app | `belt app run google/veo-3-1-fast --input input.json` |

This command runs a cloud app with an input file, which can send user data to external services. The command is expected but has real external execution impact.

Medium

SKILL.md:87-88

Ruby/shell backtick execution

| Run without waiting | `belt app run <app> --input input.json --no-wait` |

The no-wait command submits work to a cloud app and returns before completion. That external side effect is real even though the syntax is documented and bounded.

Medium

SKILL.md:104-130

Ruby/shell backtick execution

```bash

The related-skills block instructs users to install multiple additional skills with npx. This can fetch new third-party instructions and broaden the trusted code surface.

Medium

SKILL.md:28

Shell command substitution

> curl -LO $(curl -fsSL https://dist.inference.sh/cli/manifest.json | grep -o '"url":"[^"]*"' | grep

The command substitution builds a download URL from remote manifest output and passes it into curl. If the manifest or parsing is compromised, the selected artifact can be controlled remotely.

Medium

SKILL.md:25-34

Template literal with command substitution

> ```bash

The manual install block contains command substitution and remote downloads. It is documented, but it remains a supply-chain shell pattern that should be reviewed before use.

Medium

SKILL.md:33

Hidden file access

> mv inferencesh-cli-* ~/.local/bin/inferencesh

The command writes into a hidden home-directory path as part of installing the CLI binary. This is expected for local installs but remains a filesystem persistence risk.

Low

references/authentication.md:6

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

The URL is used in a command that downloads an installer and pipes it to sh. The domain appears intended, but remote installer execution creates supply-chain risk.

Low

references/authentication.md:44

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

The reinstall command fetches the same remote installer URL before shell execution. This is an expected endpoint but still a network-based installation risk.

Low

references/cli-reference.md:6

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

The URL is part of a remote installer command that is intended to execute locally. The network endpoint is expected, but it still controls the fetched install script.

Low

SKILL.md:18

Hardcoded URL

curl -fsSL https://cli.inference.sh | sh

The URL is embedded in the pipe-to-shell installer command. The external host controls the script content that would be executed locally.

Low

SKILL.md:22

Hardcoded URL

> **What does the installer do?** The [install script](https://cli.inference.sh) detects your OS and

The URL points to the installer script in explanatory text. The link is legitimate context, but it is security-relevant because the same endpoint is used for local installation.

Low

SKILL.md:27

Hardcoded URL

> curl -LO https://dist.inference.sh/cli/checksums.txt

The command downloads a checksum file from the distribution host as part of manual installation. This is expected network access but still part of the binary supply chain.

Low

SKILL.md:28

Hardcoded URL

> curl -LO $(curl -fsSL https://dist.inference.sh/cli/manifest.json | grep -o '"url":"[^"]*"' | grep

The nested curl retrieves a manifest from the distribution host and uses it to choose an artifact URL. That remote dependency affects which binary is downloaded.

Risk Factors

🌐 Network access (17)

references/authentication.md:6 references/authentication.md:44 references/cli-reference.md:6 references/running-apps.md:152 SKILL.md:9 SKILL.md:13 SKILL.md:18 SKILL.md:22 SKILL.md:27 SKILL.md:28 SKILL.md:141 SKILL.md:142 SKILL.md:143 SKILL.md:144 SKILL.md:145 SKILL.md:146 SKILL.md:147

🔑 Env variables (3)

references/authentication.md:30 references/authentication.md:53 references/cli-reference.md:79

📁 Filesystem access (5)

references/cli-reference.md:91 references/cli-reference.md:91 references/running-apps.md:41 SKILL.md:33 SKILL.md:33

⚙️ External commands (22)

Detected Patterns

Pipe to shell pattern×4

Audited by: codex View Audit History →

Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report

Copy report link

https://skillstore.io/skills/qu-skills-infsh-cli/audits/5?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/qu-skills-infsh-cli/security.svg)](https://skillstore.io/skills/qu-skills-infsh-cli?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/qu-skills-infsh-cli?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/qu-skills-infsh-cli/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/qu-skills-infsh-cli.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>

Academic citations (APA · BibTeX · CFF)

APA citation

qu-skills. (2026). infsh-cli security audit report (audit version 5) [Author version unspecified]. Skillstore. https://skillstore.io/skills/qu-skills-infsh-cli/audits/5

BibTeX citation

@techreport{qu-skills-qu-skills-infsh-cli-2026,
  author = {qu-skills},
  title = {infsh-cli security audit report (audit version 5)},
  institution = {Skillstore},
  year = {2026},
  number = {5},
  url = {https://skillstore.io/skills/qu-skills-infsh-cli/audits/5},
  note = {Author version unspecified}
}

CITATION.cff

cff-version: 1.2.0
message: "If you use this Skill, cite its author and this versioned security audit report."
title: "infsh-cli security audit report (audit version 5)"
version: "unspecified"
type: report
authors:
  - name: "qu-skills"
date-released: "2026-07-09"
url: "https://skillstore.io/skills/qu-skills-infsh-cli/audits/5"
identifiers:
  - type: other
    value: "skillstore:qu-skills-infsh-cli:audit:5"
    description: "Skillstore immutable audit report identifier"

Compare variants

9 installable variants

Each author remains a separate installable skill. The recommended variant is ranked by Skillstore evidence.

Why this variant is first

Higher Skillstore usage

Author / source Skillstore Score Evidence Confidence Skillstore usage Updated

inferen-sh Recommended

inferen-sh-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 41

Updated

2026-08-21

qu-skills Current

qu-skills-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 19

Updated

2026-08-21

tool-belt

tool-belt-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 13

Updated

2026-08-21

skillssh

skillssh-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-08-21

inference-skills

inference-skills-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 9

Updated

2026-08-21

inference-sh-skills

inference-sh-skills-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 8

Updated

2026-08-21

infsh-skills

infsh-skills-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 7

Updated

2026-08-21

101-skills

101-skills-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 5

Updated

2026-08-21

halt-catch-fire

halt-catch-fire-infsh-cli

Skillstore Score 38

Evidence Confidence Medium

Skillstore usage 3

Updated

2026-08-21

Skillstore Score

Why this score Evidence Confidence: Medium

Architecture

Maintainability

Content

Community

Spec Compliance

What You Can Build

Generate media assets

Create images, videos, audio, or 3D assets from prompts using hosted inference.sh apps.

Run model experiments

Compare LLM, search, and media apps from one CLI during prototyping.

Automate cloud AI jobs

Submit long-running jobs, check task status, and save generated results for workflows.

Try These Prompts

Find an app

Find inference.sh apps for image generation and explain which one fits a simple product mockup task.

Prepare a run

Help me inspect an inference.sh app, create a sample input file, and run it with a short prompt.

Use local media

Run an inference.sh app with a local image file. Show me the files that will be uploaded before execution.

Build a tracked workflow

Design a belt CLI workflow that submits a video generation job, returns the task ID, checks progress, and saves the final output.

Best Practices

Confirm account, cost, and provider terms before running paid or third-party apps.
Review every local file path before upload, especially private media or customer data.
Require explicit approval before any Twitter/X posting, direct messaging, following, liking, or reposting.

Avoid

Do not pipe remote install scripts to a shell without reviewing and verifying them.
Do not place API keys, private prompts, or customer data directly in shared command history.
Do not run social media write actions without a clear user request and final confirmation.